The basic concept to be investigated is how to adapt biology to being rationally programmable, so that molecules and eventually organisms can be more readily engineered for a variety of research and commercial applications. Extensive genetic programmability will require a new type of biological computer. The term "amorphous computer" was coined by MIT researchers to describe computing systems comprised of very large numbers of identical computers each of which possessed limited processing power, limited memory, local communcation, no a priori knowledge of position, and no synchronizing clock. The description as "amorphous" is apropos - the results of algorithms executing on such computers emerge from a shapeless, seemingly unorganized, mass. In order to implement practical, amorphous computations, we have modeled and are beginning to build two-dimensional arrays of transcriptional logic gates. In practice, RNA molecules transcribed from one promoter diffuse, bind to another promoter, and either activate it or inactivate it. The advantages of such transcriptional logic gates is that the address space is essentially as large as nucleic acid sequence space (scaling to 4n). Milestones that will build towards a generalized platform for amorphous computation include: 1. Developing a reproducible testbed for programming on surfaces. 2. Pattern generation from immobilized 'toggle'switches. 3. Generating a programmed behavior: the stadium wave. 4. Signal amplification and sensor function. By developing algorithms that rely upon diffusible, information rich molecules to actuate gate structures we are beginning to build biological computers that run modular genetic software. The principles that we acquire in developing this software will have an impact well beyond any individual algorithms or instantiations. The 2-D arrays and accompanying molecular computations will become a testbed for both modeling and experimenting with reaction-diffusion kinetics in complex informational systems. Beyond enabling amplification of signals from molecular sensors (Milestone 4), these experiments also address one of the key problems in nanotechnology: how to program the self-assembly of complex devices. PUBLIC HEALTH RELEVANCE: We propose to develop a new type of molecular computer, an amorphous computer. This computer will operate much like organisms do: individual processors (like cells) will be programmed to carry out a limited set of operations (like eating sugar) based on diffusible signals (like the hormone, insulin). However, instead of cells we will use DNA elements as the processors. The DNA elements will make diffusible RNA molecules that will move between, and alter the state and function of, the processors. We suggest a graded approach to the construction of our new type of computer, building from a standardized testbed through a static demonstration of pattern formation to a dynamic demonstration of pattern formation to the application as a sensor for an important protein in the blood, platelet-derived growth factor. The new genetic computer that we develop will be modular and expandable, and will create a new paradigm that allows for the rational development of biological software. The results of these inquiries should help understand development, including how development sometimes goes awry during disease formation, and may assist with building nanoscale devices for therapy and diagnostics.